A three-step support strategy for relatives of patients dying in the intensive care unit: a cluster randomised trial.

BACKGROUND: In relatives of patients dying in intensive care units (ICUs), inadequate team support can increase the prevalence of prolonged grief and other psychological harm. We aimed to evaluate whether a proactive communication and support intervention would improve relatives' outcomes. METHODS: We undertook a prospective, multicentre, cluster randomised controlled trial in 34 ICUs in France, to compare standard care with a physician-driven, nurse-aided, three-step support strategy for families throughout the dying process, following a decision to withdraw or withhold life support. Inclusion criteria were relatives of patients older than 18 years with an ICU length of stay 2 days or longer. Participating ICUs were randomly assigned (1:1 ratio) into an intervention cluster and a control cluster. The randomisation scheme was generated centrally by a statistician not otherwise involved in the study, using permutation blocks of non-released size. In the intervention group, three meetings were held with relatives: a family conference to prepare the relatives for the imminent death, an ICU-room visit to provide active support, and a meeting after the patient's death to offer condolences and closure. ICUs randomly assigned to the control group applied their best standard of care in terms of support and communication with relatives of dying patients. The primary endpoint was the proportion of relatives with prolonged grief (measured with PG-13, score ≥30) 6 months after the death. Analysis was by intention to treat, with the bereaved relatives as the unit of observation. The study is registered with ClinicalTrials.gov, NCT02955992. FINDINGS: Between Feb 23, 2017, and Oct 8, 2019, we enrolled 484 relatives of ICU patients to the intervention group and 391 to the control group. 379 (78%) relatives in the intervention group and 309 (79%) in the control group completed the 6-month interview to measure the primary endpoint. The intervention significantly reduced the number of relatives with prolonged grief symptoms (66 [21%] vs 57 [15%]; p=0·035) and the median PG-13 score was significantly lower in the intervention group than in the control group (19 [IQR 14-26] vs 21 [15-29], mean difference 2·5, 95% CI 1·04-3·95). INTERPRETATION: Among relatives of patients dying in the ICU, a physician-driven, nurse-aided, three-step support strategy significantly reduced prolonged grief symptoms. FUNDING: French Ministry of Health.

ramR Deletion in an Enterobacter hormaechei Isolate as a Consequence of Therapeutic Failure of Key Antibiotics in a Long-Term Hospitalized Patient.

Genome changes are central to the adaptation of bacteria, especially under antibiotic pressure. The aim of this study was to report phenotypic and genomic adaptations undergone by an Enterobacter hormaechei clinical strain that became highly resistant to key antimicrobials during a 4-month period in a patient hospitalized in an intensive care unit (ICU). All six clinical E. hormaechei strains isolated in one ICU-hospitalized patient have been studied. MICs regarding 17 antimicrobial molecules have been measured. Single nucleotide polymorphisms (SNPs) were determined on the sequenced genomes. The expression of genes involved in antibiotic resistance among Enterobacter cloacae complex strains were determined by reverse transcription-quantitative PCR (qRT-PCR). All the strains belonged to sequence type 66 and were distant by a maximum of nine SNPs. After 3 months of hospitalization, three strains presented a significant increase in MICs for ceftazidime, cefepime, temocillin, ertapenem, tigecycline, ciprofloxacin, and chloramphenicol. Those resistant strains did not acquire additional antibiotic resistance genes but harbored a 16-bp deletion in the ramR gene. This deletion led to upregulated expression of RamA, AcrA, AcrB, and TolC and downregulated expression of OmpF. The ΔramR mutant harbored the same phenotype as the resistant clinical strains regarding tigecycline, chloramphenicol, and ciprofloxacin. The increased expression of RamA due to partial deletion in the ramR gene led to a cross-resistance phenotype by an increase of antibiotic efflux through the AcrAB-TolC pump and a decrease of antibiotic permeability by porin OmpF. ramR appears to be an important adaptative trait for E. hormaechei strains.

Human neutrophils produce interferon gamma upon stimulation by interleukin-12.

Interferon gamma (IFNgamma) is a Th1 cytokine mainly produced by T cells, NK cells and macrophages in response to interleukin (IL)-12. As polymorphonuclear neutrophils (PMN) have been shown to produce and to release numerous cytokines, in particular upon IL-12 stimulation, we investigated the ability of highly purified PMN to secrete IFNgamma. We found that PMN contained a small store of IFNgamma, and that this store was rapidly secreted upon stimulation by degranulating agents such as formyl peptides. Moreover, after a few hours of stimulation with appropriate agents, PMN synthesized IFNgamma. The effect of IL-12 was time- and concentration-dependent, and IL-12 combinations with IL-2, IL-15, IL-18 or LPS were highly synergistic. Cycloheximide inhibited IFNgamma release in such optimal conditions, confirming the ability of PMN to synthesize IFNgamma. IFNgamma synthesis was associated with an increase in specific mRNA content, pointing to a transcriptional mechanism. The IFNgamma produced by PMN was biologically active, as demonstrated by its ability to induce TNFalpha synthesis by PMN themselves or to induce IL-10 synthesis by peripheral blood mononuclear cells. These findings reveal a novel pathway of autocrine and paracrine PMN activation. They also identified a new role for IFNgamma, bridging innate and adaptive immune responses.

Regulation of interleukin 12 p40 and p70 production by blood and alveolar phagocytes during severe sepsis.

Paradoxically, the host response to severe sepsis may lead to immunosuppression, thereby favoring nosocomial infections. We examined the role of the two IL-12 isoforms, bioactive IL-12p70 and regulatory IL-12p40, in 16 patients with severe sepsis. We compared the capacity of purified blood and alveolar phagocytes [polymorphonuclear neutrophils (PMN) and monocytes/macrophages] to secrete each isoform. Blood monocytes had normal basal secretions. In contrast, a marked imbalance was observed after ex vivo stimulation by lipopolysaccharide plus IFN-gamma, with significantly lower IL-12p70 production and higher IL-12p40 production. Conversely, stimulated IL-12p40 production by the patients' blood PMN tended to be impaired, as was their cell-surface beta2 integrin and L-selectin expression, known as markers of cell activation. In the patient's bronchoalveolar lavage fluid, the production of both IL-12 isoforms after ex vivo stimulation was significantly lower with alveolar macrophages than with autologous blood monocytes and significantly higher with alveolar PMN than with autologous blood PMN. This sheds new light on the potential role of PMN in local modulation of inflammation, via secretion of the anti-inflammatory IL-12 p40 subunit. The imbalance between the bioactive and regulatory IL-12 isoforms, which is probably designed to control excessive inflammation, may also make septic patients more susceptible to nosocomial infection.

Pharmacokinetics of long-term sufentanil infusion for sedation in ICU patients.

OBJECTIVE: To determine the pharmacokinetics of long-term infusion of sufentanil in ICU patients. DESIGN AND SETTING: Open-label study in a surgical intensive care unit. PATIENTS: Ten consecutive patients without renal or hepatic failure requiring mechanical ventilation for at least 6 days. INTERVENTIONS: Patients received sufentanil (initial bolus 0.5 micro g/kg and continuous infusion rate of 0.5 micro g/kg per hour) and midazolam (initial bolus 0.08 mg/kg and continuous infusion 0.05 mg/kg per hour). Sedation was adjusted according to the Ramsay scale (score >3). Blood samples were taken during and up to 72 h after the infusion, and plasma concentrations were measured using a sensitive radioimmunoassay method. MEASUREMENTS AND RESULTS: Plasma concentration-time profiles of sufentanil and pharmacokinetic parameters such as initial postinfusion half-life (t(1/2alpha)), elimination half-life (t(1/2beta)), total clearance (Cl), volume of distribution (Vdbeta), and time required to obtain a 50% decrease in plasma concentration (tcp(0/2)). The mean duration of sedation was 12+/-7 days. The initial half-life t(1/2alpha) was 1.33+/-1.15 h. The observed prolonged elimination half-life (t(1/2beta)=25.5+/-9.4 h) was related to the large volume of distribution (Vdbeta=22.6+/-9.4 l/kg). The mean total clearance was 13.4+/-7.0 ml/kg per minute. The mean time required to obtain a 50% decrease in plasma concentration was short (tcp(0/2=)4.7+/-3.7 h). CONCLUSIONS: The pharmacokinetic analysis of sufentanil for ICU sedation revealed increased volume of distribution and elimination half-life. Nevertheless the rapid distribution and elimination processes suggest that the rapid reversibility of sedation with sufentanil is maintained after long duration of infusion. Further studies should be carried out to evaluate the clinical relevance of these results.

Ethanol-induced inhibition of cytokine release and protein degranulation in human neutrophils.

Ethanol impairs immune responses in humans and animal models, in vivo and in vitro. In particular, ethanol inhibits some key functions of human polymorphonuclear neutrophils (PMN). We investigated the impact of ethanol on cytokine production by highly purified PMN. In a time- and concentration-dependent manner, ethanol inhibited the production of interleukin (IL)-8 protein and mRNA and also hindered tumor necrosis factor alpha (TNF-alpha) release by modulating the expression of the TNF-alpha-converting enzyme involved in TNF-alpha shedding. This disruption of PMN cytokine release by ethanol may contribute to the increased risk of infection in alcoholic patients. Degranulation of hepatocyte growth factor (HGF) was also impaired by a clinically relevant ethanol concentration (0.8%), an action that may delay the repair of alcoholic liver damage. These findings suggest that ethanol may modulate three major cytokines involved in alcoholic liver diseases, IL-8, TNF-alpha, and HGF, via three different mechanisms.